Deficits in attention are prominent in clinical studies of cocaine users, and likely contribute to a broad range of cognitive deficits associated with cocaine dependence. Understanding the basis of these cognitive deficits is clinically important because of the demonstrated relationship between cognitive performance and treatment outcome. Two distributed cortical networks interact in an anti- correlated fashion in support of focused attentional processing. A dorsal frontoparietal system becomes activated during purposeful cognitive tasks. Another distributed network, the default mode network deactivates from a high metabolic activity at rest to a lower level during focused cognitive tasks. Impaired deactivation of that default mode network is associated with attentional lapses and clinical populations with attentional deficits show impaired integrity and phasic deactivation of the default mode network. Dopamine is implicated in attentional processing, and recent evidence demonstrates an inverse correlation between the dopamine transporter and default network deactivation. There is also evidence for inadequate dopaminergic neurotransmission in human cocaine users during short term withdrawal, in part from an up-regulated dopamine transporter. In the present application, we propose to examine the overarching hypothesis that impaired attention following chronic cocaine use is associated with inadequate deactivation of the default mode network during focused attention, and that insufficient dopaminergic neurotransmission is a proximate cause of the impaired default network deactivation. We will investigate this hypothesis by accomplishing four aims: Aim 1) Use an attentional task with parametrically variable attentional demand to establish matched performance for functional brain imaging studies in a control, and a chronic cocaine self-administering group of monkeys with impaired attention. Aim 2) Quantify the dopamine transporter in the two groups of animals using high resolution microPET imaging. Aim 3) Determine if the chronic cocaine group shows impaired default mode network deactivation during attentional processing using microPET measures of regional glucose metabolism. Aim 4) Determine if treatment with a low dose of the indirect dopamine agonist amphetamine improves attentional control concomitant with improved deactivation of the default mode network measured by CMRglc.